Journal of reproductive immunology · 2026
Development and validation of cell-free RNA-based predictive models for preeclampsia using placental single-cell transcriptomic data.
СтатьяВалидационное исследованиеНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Chen Yinlong, Zhan Wenli, Shen Ke, Chen Zhi, Zheng Qingliang
Аннотация
INTRODUCTION: Preeclampsia (PE) is a serious complication of pregnancy for which reliable early predictive biomarkers are currently lacking. Cell-free RNA (cfRNA) has emerged as a promising non-invasive approach that may enhance clinical management and facilitate earlier diagnosis.
METHODS: To investigate PE, we applied single‑cell RNA sequencing (scRNA‑seq) to placental tissues, enabling the identification of differentially expressed genes. Integration of weighted gene co‑expression network analysis (WGCNA) with cfRNA profiles from PE patients revealed two gene modules that showed significant relevance to the pathogenesis of PE. By intersecting the genes within these modules and applying advanced machine learning algorithms, we successfully pinpointed optimal cfRNA biomarkers for PE prediction. Furthermore, we constructed a nomogram and predictive model to enhance clinical utility, complemented by immune infiltration analysis and multi-database expression validation to elucidate the functional relevance of these biomarkers in PE.
RESULTS: A diagnostic nomogram was established based on five candidate cfRNAs: KRT19, CCND1, ZBTB16, TUBB6, and HMGB2. In the training cohort, the model yielded strong performance with an AUC of 0.927 (95% CI: 0.872-0.981). Subsequent validation across two independent datasets confirmed its robustness, producing AUCs of 0.910 (95% CI: 0.806-1.000) and 0.772 (95% CI: 0.714-0.830), respectively.
CONCLUSIONS: A candidate PE predictive model was constructed from placental single-cell transcriptomic data with a five-cfRNA panel, warranting further large multi-centre prospective validation.
METHODS: To investigate PE, we applied single‑cell RNA sequencing (scRNA‑seq) to placental tissues, enabling the identification of differentially expressed genes. Integration of weighted gene co‑expression network analysis (WGCNA) with cfRNA profiles from PE patients revealed two gene modules that showed significant relevance to the pathogenesis of PE. By intersecting the genes within these modules and applying advanced machine learning algorithms, we successfully pinpointed optimal cfRNA biomarkers for PE prediction. Furthermore, we constructed a nomogram and predictive model to enhance clinical utility, complemented by immune infiltration analysis and multi-database expression validation to elucidate the functional relevance of these biomarkers in PE.
RESULTS: A diagnostic nomogram was established based on five candidate cfRNAs: KRT19, CCND1, ZBTB16, TUBB6, and HMGB2. In the training cohort, the model yielded strong performance with an AUC of 0.927 (95% CI: 0.872-0.981). Subsequent validation across two independent datasets confirmed its robustness, producing AUCs of 0.910 (95% CI: 0.806-1.000) and 0.772 (95% CI: 0.714-0.830), respectively.
CONCLUSIONS: A candidate PE predictive model was constructed from placental single-cell transcriptomic data with a five-cfRNA panel, warranting further large multi-centre prospective validation.